Frontal tau pathology underlies behavioural / dysexecutive clinical presentations of Alzheimer’s disease

Abstract

AbstractBackgroundIncreasing attention is being directed to atypical clinical presentations of Alzheimer’s disease. The behavioural/dysexecutive variant of Alzheimer’s disease is a rare clinical syndrome presenting with behavioural changes, apathy and executive dysfunction, similar to frontotemporal dementia. While initial autopsy studies identified pathology in the frontal cortex, recent studies challenge the notion of frontal involvement in this condition.MethodWe assessed 14 cases of behavioural/dysexecutive Alzheimer’s disease recruited from a tertiary care memory clinic, all of whom had biologically defined Alzheimer’s disease (amyloid and tau positive). They were compared with 24 disease severity‐ and age‐matched amnestic Alzheimer’s disease patients and a group of 131 cognitively unimpaired (CU) elderly individuals. All subjects were evaluated with [18F]AZD4694‐PET (amyloid‐b), [18F]MK6240‐PET (tau), MRI, genotyping for apolipoprotein E (APOE) and neuropsychological testing.ResultVoxelwise contrasts against CU elderly identified patterns of frontal cortical tau aggregation in the behavioural/dysexecutive group, with peaks in the anterior cingulate, frontal insula and orbitofrontal cortices. Amnestic Alzheimer’s disease patients had larger differences in the lateral temporal, inferior parietal and posterior cingulate cortices. In contrast to tau‐PET, no differences were observed in the distribution of amyloid‐PET between both Alzheimer’s disease groups. Finally, voxelwise regressions demonstrated significant relationships between frontal cortical tau load and severity of executive dysfunction (FWE corrected at p<0.001), while memory dysfunction was related to temporal and parietal tau load (FWE corrected at p<0.001).ConclusionThe association between frontal tau pathology and behavioural / dysexecutive clinical presentations provides in vivo support for a “frontal variant” of Alzheimer’s disease as initially identified by autopsy studies. Our results also highlight the need for consensus clinical criteria for behavioural/dysexecutive Alzheimer’s disease.